ARTICLE

Prostaglandin E2/EP2 receptor signalling pathway promotes diabeticretinopathy in a rat model of diabetes

Man Wang1& Yangningzhi Wang1

& Tianhua Xie1& Pengfei Zhan1

& Jian Zou2& Xiaowei Nie2,3

& Jun Shao1&

Miao Zhuang1& Chengye Tan1

& Jianxin Tan2& Youai Dai2 & Jie Sun2

& Jiantao Li4 & Yuehua Li4 & Qian Shi5 & Jing Leng6&

Xiaolu Wang2,3& Yong Yao1

Received: 4 May 2018 /Accepted: 20 September 2018 /Published online: 8 November 2018# Springer-Verlag GmbH Germany, part of Springer Nature 2018

AbstractAims/hypothesis Diabetic retinopathy is a common microvascular complication of diabetes mellitus and is initiated by inflam-mation and apoptosis-associated retinal endothelial cell damage. Prostaglandin E2 (PGE2) has emerged as a critical regulator ofthese biological processes. We hypothesised that modulating PGE2 and its E-prostanoid receptor (EP2R) would prevent diabetesmellitus-induced inflammation and microvascular dysfunction.Methods In a streptozotocin (STZ)-induced rat model of diabetes, rats received intravitreal injection of PGE2, butaprost (a PGE2/EP2R agonist) or AH6809 (an EP2R antagonist). Retinal histology, optical coherence tomography, ultrastructure of the retinalvascular and biochemical markers were assessed.Results Intravitreal injection of PGE2 and butaprost significantly accelerated retinal vascular leakage, leucostasis and endothelialcell apoptosis in STZ-induced diabetic rats. This response was ameliorated in diabetic rats pre-treated with AH6809. In addition,pre-treatment of human retinal microvascular endothelial cells with AH6809 attenuated PGE2- and butaprost-induced activationof caspase 1, activation of the complex containing nucleotide-binding domain and leucine rich repeat containing family, pyrindomain containing 3 (NLRP3) and apoptosis-associated speck-like protein containing a C-terminal caspase-activation andrecruitment domain (ASC), and activation of the EP2R-coupled cAMP/protein kinase A/cAMP response element-binding proteinsignalling pathway.Conclusions/interpretation The PGE2/EP2R signalling pathway is involved in STZ-induced diabetic retinopathy and could beconsidered as a potential target for diabetic retinopathy prevention and treatment.

Keywords Diabetic retinopathy . Endothelial cell . Inflammation .Microvascular disease . PGE2

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s00125-018-4755-3) contains peer-reviewed butunedited supplementary material, which is available to authorised users.

* Yong Yaoyongyao@njmu.edu.cn

* Xiaolu Wangxlwang@njmu.edu.cn

1 Department of Ophthalmology, Wuxi People’s Hospital Affiliated toNanjing Medical University, 299 Qingyang Road, Wuxi, Jiangsu,People’s Republic of China

2 Center of Clinical Research, Wuxi People’s Hospital Affiliated toNanjing Medical University, 299 Qingyang Road, Wuxi, Jiangsu,People’s Republic of China

3 Wuxi Institute of Translational Medicine, Wuxi, Jiangsu, People’sRepublic of China

4 Key Laboratory of Cardiovascular Disease and MolecularIntervention, Department of Pathophysiology, Nanjing MedicalUniversity, Nanjing, Jiangsu, People’s Republic of China

5 Yixing Eye Hospital, Wuxi, Jiangsu, People’s Republic of China

6 Cancer Center, Department of Pathology, Nanjing MedicalUniversity, Nanjing, Jiangsu, People’s Republic of China

Diabetologia (2019) 62:335–348https://doi.org/10.1007/s00125-018-4755-3

AbbreviationsASC Apoptosis-associated speck-like protein contain-

ing a C-terminal caspase-activation and recruit-ment domain

COX-2 Cyclooxygenase-2CREB cAMP response element-binding proteinEP1–4R E-prostanoid1–4 receptorhRMEC Human RMECLDH Lactate dehydrogenaseNLRP3 Nucleotide-binding domain and leucine rich repeat

containing family, pyrin domain containing 3OCT Optical coherence tomographyPGE2 Prostaglandin E2

PKA Protein kinase ARMEC Retinal microvascular endothelial cellSTZ Streptozotocinz-VAD Z-YVAD-fmk

Introduction

Diabetic retinopathy, a prevalent complication of diabetes, is aleading cause of visual impairment and blindness in the adultpopulation. However, the biochemical and molecular mecha-nisms are not well understood [1, 2]. Early clinical symptomsinclude retinal microvascular endothelial cell (RMEC) dys-function and vascular dysfunction [3, 4]. Emerging evidenceindicates that high-glucose-induced para-inflammation,characterised by a chronic low level of inflammation and adisordered immune response, is involved in the onset and

progression of RMEC damage [5, 6]. The retinas from animalmodels of diabetes show inappropriate activation of thenucleotide-binding domain and leucine rich repeat containingfamily, pyrin domain containing 3 (NLRP3) inflammasome,which is a molecular complex of NLRP3, apoptosis-associated speck-like protein containing a C-terminal cas-pase-activation and recruitment domain (ASC) and pro-caspase 1 [7]. These diabetic retinas also express high levelsof proinflammatory cytokines such as IL-1β [8].

Two signalling pathways are believed to be involved in thegeneration and release of IL-1β induced by high glucoselevels and the accumulation of advanced glycation end-products [9]. The first pathway is triggered by glucose abnor-malities, which induce IL-1β transcription and stimulate theproduction of IL-1β precursor pro-IL-1β. The second signal-ling pathway induces conformational changes in the NLRP3inflammasome platform and activates caspase 1 to convertpro-IL-1β into the mature secreted form of IL-1β. TheNLRP3 inhibitor MCC950 has been shown to inhibit high-glucose-induced RMEC dysfunction, consistent with thepromising clinical effects of an IL-1 receptor antagonist forthe treatment of diabetic retinopathy [5].

Prostaglandin E2 (PGE2) is a potent inflammatory mediatorthat is a crucial IL-1β inducer and causes fever [10]. PGE2 isbiosynthesised from arachidonic acid by cyclooxygenase en-zyme and stimulates its G-protein-coupled plasma membranereceptors (E-prostanoid1–4 receptors [EP1–4Rs]), activatingmultiple signal transduction pathways leading to downstreamresponses [11]. The EP1 receptor mainly couples to the Gqprotein and upregulates the level of intracellular calcium; the

336 Diabetologia (2019) 62:335–348

EP2 and EP4 receptors couple to the Gs protein, activate ade-nylate cyclase and increase the production of intracellularcAMP. In contrast, the EP3 receptor couples to the Gi protein,inactivates adenylate cyclase and decreases the formation ofintracellular cAMP [12].

Recent research findings have verified that cyclooxygenase-2 (COX-2) and PGE2 are involved in the pathogenesis of dia-betic retinopathy. Significantly higher than normal PGE2 levelshave been detected in the vitreous fluid of individuals withcomplications from proliferative diabetic retinopathy and alsoin animal models of diabetic retinopathy [13]. In addition, pro-gression of retinopathy can be prevented or delayed by prosta-glandin inhibitors [14, 15]. In receptor combination patterns,PGE2 shows various biological effects and the specific E-prostanoid receptors of PGE2 that regulate endothelium impair-ment and vascular dysfunction have not been well illustrated.These findings prompted us to determine whether the PGE2/EP2R cascade mediates RMEC damage in diabetic retinopathyand to investigate the underlying molecular mechanisms.

Methods

Human vitreous fluid Participants with type 1 diabetes whohad undergone vitrectomy owing to proliferative diabetic ret-inopathy were recruited from Wuxi People’s HospitalAffiliated to Nanjing Medical University, Wuxi, Jiangsu,China. The research followed the tenets of the Declarationof Helsinki. The protocol for sample collection was approvedby the hospital ethics committee and the study participantsgave informed consent. For further details, see electronic sup-plementary materials (ESM) Methods.

Cell culture Human RMECs (hRMECs) were obtained fromBeNa Culture Collection (Beina Chuanglian BiotechnologyInstitute, Beijing, China) and cultured in DMEM supplement-ed with 10% FBS (vol./vol.) and 1% antimycotics and antibi-otics (vol./vol.). Mycoplasma contamination was not tested.For further details, see ESM Methods.

Animals and treatments Eight-week-old homozygous maleSprague Dawley rats (220–250 g) were randomly divided intosix groups. Diabetes was induced with an i.p. injection ofstreptozotocin (STZ; 60 mg/kg in 10 mmol/l citrate buffer atpH 4.6), as previously reported [16]. The rats had blood glucoselevels >16.7 mmol/l, indicating that diabetes had been success-fully established. The STZ-treated rats were given an intravit-real injection of 5 mmol/l PGE2, butaprost (a PGE2/EP2R ago-nist) or AH6809 (an EP2R antagonist), all mixed with salinesolution (154 mmol/l NaCl) 1:1, at a total volume of 6 μl foreach eye. A vehicle control was prepared by mixing one vol-ume of DMSO with one volume of saline solution. The sixexperimental groups were as follows: control; untreated STZ;

STZ + PGE2; STZ + butaprost; STZ + AH6809 and STZ +DMSO. All studies adhered to the institutional guidelines forhumane treatment of animals, Principles of Laboratory AnimalCare (National Institutes of Health [NIH], Bethesda, MD,USA) and to the Association for Research in Vision andOphthalmology (ARVO) Statement for the Use of Animals inOphthalmic and Vision Research. For further details, see ESMMethods.

Intravitreal injection Rats were ventilated after beinganaesthetised with a mixture of ketamine (80 mg/kg, i.p.)and xylazine (4 mg/kg, i.p.). A volume of about 6 μl of thedesignated mixture was delivered into the vitreous cavityusing a 33-gauge needle. Rats received an intravitreal injec-tion every 3 weeks.

Retinal imaging Rats were anaesthetised (ketamine/xylazine)and their pupils were dilated with Cyclomydril (Alcon, FortWorth, TX, USA). Spectral domain optical coherence tomog-raphy (OCT) was performed using the image-guided OCTsystem (Micron IV; Phoenix Research Labs, Pleasanton,CA, USA) with the guidance of a bright-field live fundusimage.

Permeability measurement The permeability of the blood–retina barrier in rats was quantified with Evans Blue, whichbinds to the plasma albumin, using the method described byShan and colleagues [16] (see ESMMethods). Digital imagesof rat retinal flat mounts were examined under an OlympusBX-51 light microscope (Olympus, Tokyo, Japan) to checkfor Evans Blue extravasation from the retinal vessels.

Retinal trypsin digestion assay Eyes of rats were enucleated,fixed in 4% paraformaldehyde (wt/vol.) for 24 h, equatoriallybisected and the retinas were removed. The retinas were incu-bated with 3% trypsin (wt/vol.) at 37°C for 3 h; they were thengently shaken to free the vessel network, washed and mountedonto glass slides to dry. Retinal vasculature was stained withPeriodic acid–Schiff and Haematoxylin. Digital images wereexamined under an Olympus BX-51 light microscope(Olympus, Tokyo, Japan).

H&E staining Eyes of rats were enucleated and fixed in 4%paraformaldehyde (wt/vol.) for 24 h. The retina and sclerawere dehydrated in a graded ethanol series and embedded inparaffin. For H&E staining, 5 μm thick sections were takenalong the vertical meridian and observed under an OlympusBX-51 light microscope (Olympus, Tokyo, Japan).

Immunofluorescence analysis Standard immunofluorescenceanalysis was performed to localise NLRP3 andASC expressionin retina of rats as previously described [17]. For TUNEL anal-ysis, the retina sections of rats were stained using a fluorescein-

Diabetologia (2019) 62:335–348 337

conjugated TUNEL in situ cell death detection kit (RocheDiagnostics, Mannheim, Germany). Images were acquiredusing a confocal microscope (Leica, Heidelberg, Germany).

Lectin labelling of the adherent retinal leucocytes Leucostasisassay was performed using a perfusion labelling technique, aspreviously reported [18] (see ESM Methods). FITC-coupledconcanavalin A lectin (Vector Labs, Burlingame, CA, USA)was used to label the adherent leucocytes and the vascularendothelial cells in rat retinas. Images were examined undera confocal microscope (Leica, Heidelberg, Germany).

Transmission electron microscopy analysis Retinal sections ofrats, approximately 2 mm× 3 mm, were isolated from eacheyecup following the protocol described previously [19] (seeESM Methods). The ultrastructure of the retina tissues wasobserved using a transmission electron microscope (TecnaiG2 Spirit Bio TWIN; FEI, Hillsboro, OR, USA).

Immunoprec ip i tat ion and western blot analys isImmunoprecipitation and western blotting were performedas described previously [17] using NLRP3, ASC, caspase 1,caspase 3, CREB, p-CREB, EP1R, EP2R, EP3R, EP4R,COX-2, Epac1, β-actin and Laminb antibodies (see ESMMethods).

RNA quantification The relative expression levels of mRNAof rat Icam1, rat Il-1β, human IL-1β, human NLRP3, humanEP1R, human EP2R, human EP3R or human EP4R werequantified by quantitative RT-PCR [17] (see ESM Methods).The data were analysed by using the 2−ΔΔCt method and nor-malised to endogenous control GAPDH or Gapdh mRNA.The primers used are detailed in ESM Table 1.

Tissue and serum biochemical measurements Serum lactatedehydrogenase (LDH) levels in the cell culture supernatantfractions of the hRMECs were measured using commerciallyavailable assays (Nanjing Jiancheng Bioengineering Institute,Nanjing, China).

Flow cytometry The hRMECs were suspended in 400 μl ofbinding buffer (422201, Biolegend, San Diego, CA, USA)and stained with 5 μl of annexin V–FITC at 4°C in the dark.After 15min, the cells were incubated with 10μl of propidiumiodide buffer for 5 min at 4°C in the dark. The cell apoptosisrates were evaluated by a Cytomics FC500 instrument(Beckman Coulter, Miami, FL, USA).

IL-1β and PGE2 assay IL-1β and PGE2 in the cell culturesupernatant fractions of the hRMECs was measured usingcommercial ELISA kits (Elabscience Biotechnology Co.,Wuhan, China; Cayman Chemical Company, Ann Arbor,MI, USA, respectively). The cell culture supernatant fraction

was concentrated tenfold by ultrafiltration centrifugation(Amicon Ultra-0.5; Millipore, Billerica, MA, USA).

Statistical analysisThe experiments were not performed blind.However, an effort was made to simulate the conditions ofblinded assays. All the samples were obtained via the sameprocedures and treated in the same way. All the data wasobtained via direct recording of the physiological variablesso that the analysis did not include any subjective evaluations.The results are expressed as the mean ± SEM. Significancewas established between two groups using Student’s t test(paired t test), while ANOVA was used for multiple groupcomparisons followed by Tukey’s post hoc test. Tukey’s posthoc test was run only if the F value achieved p < 0.05 andthere was no significant variance in homogeneity. The datawas analysed with the GraphPad Prism-5 statistical software(Prism v5.0; GraphPad Software, La Jolla, CA, USA).Differences were considered statistically significant atp < 0.05.

Results

PGE2/EP2R signalling mediates the impairment of retinal ves-sels in rat model of type 1 diabetes The potent proinflamma-tory and angiogenic cytokine PGE2 was upregulated in thevitreous fluid of individuals with diabetic retinopathy vshealthy individuals (ESM Fig. 1a). We cultured hRMECs inhigh-glucose medium to mimic diabetic conditions in vitro.High-glucose stress resulted in a significant increase in COX-2 and EP2R expression compared with the control medium(ESM Fig. 1b–d). Moreover, EP2R (also known asPTGER2) mRNA levels were significantly increased inhRMECs treated with ATP and lipopolysaccharide (importantrisk factors for endothelial cell injury in diabetic retinopathy)(ESM Fig. 1e). This data suggests a potential role for EP2R inRMECs experiencing high-glucose stress.

Diabetic retinopathy is typically characterised by an abnor-mal change in the retinal microvasculature, resulting in retinalnon-perfusion, increased vasopermeability and pathologicalintraocular proliferation of the retinal vessels [20]. We usedintravitreal injection of an EP2R agonist (butaprost) and an-tagonist (AH6809) in STZ-induced diabetic rats for 3 monthsto investigate whether EP2R is a potential regulator ofdiabetes-induced microvascular complications. Evans Blueleakage assay indicated that PGE2 and butaprost increaseddiabetes-induced retinal vascular leakage; these symptomswere alleviated in rats that were pre-treated with AH6809(Fig. 1a–c). Retinal trypsin digestion assay indicated thatdiabetes-associated pericyte loss and capillary degenerationwere more severe in the retinas of PGE2- and butaprost-treated rats (Fig. 1d,e). Inhibition of EP2R partially reducedthis detrimental effect (Fig. 1d,e).

338 Diabetologia (2019) 62:335–348

PGE2/EP2R signalling influences morphological changes indiabetic rat retina The fundus images taken in PGE2- andbutaprost-treated diabetic rats showed that the intraretinalmicrovascular abnormalities occurred as early as 6 weeksafter diabetes was successfully established (Fig. 2a). Atthat time, the deposition of extravasated lipoproteins ineach group of rats was not yet clearly visible by fundusphotography. OCT showed an increased retinal thicknessin the PGE2- and butaprost-treated diabetic rats. The num-ber of hyperreflective dots (arrows) in the superficial por-tion of the inner retina dramatically increased in thePGE2- and butaprost-treated diabetic rats (Fig. 2a,b). In

contrast, the AH6809 pre-treated group showed signifi-cantly improved morphology of the retinal layers underdiabetic conditions (Fig. 2a,b). We continued to monitorthe retinal oedema and detachments later in the course ofthe disease model. Three months after diabetes was suc-cessfully established, histological examination showedthat the retinal tissue in the control rats housed undernormal conditions was intact and that the layers of theretina were clear and regularly arranged (Fig. 2c). Bycomparison, the retinal oedema in the STZ-treated ratswas remarkable, and in the PGE2- and butaprost-treatedrats, the disordered retinal structure, retinal oedema and

Con STZ STZ+PGE2

Con STZ STZ+PGE2

STZ+butaprost STZ+AH6809 STZ+DMSO

STZ+butaprost STZ+AH6809 STZ+DMSO

0.25

0.20

0.15

0.10

0.05

Eva

ns B

lue

area

(%

)E

vans

Blu

e(µ

g/m

g tis

sue)

Ace

llula

r ve

ssel

num

ber/

field

0

**

****

**

*

**

*

*

**

*

* **

ConSTZ

STZ+PGE 2

STZ+but

apro

st

STZ+AH68

09

STZ+DM

SO

ConSTZ

STZ+PGE 2

STZ+bu

tapr

ost

STZ+AH68

09

STZ+DM

SO

ConSTZ

STZ+PGE 2

STZ+bu

tapr

ost

STZ+AH68

09

STZ+DM

SO

1.0

0.8

0.6

0.4

0.2

0

20

15

10

5

0

a

d

b

c

e

Fig. 1 PGE2/EP2R signalling mediates retinal vascular leakage and cap-illary degeneration in diabetic rats. (a–c) Rats were infused with EvansBlue dye for 2 h. Red fluorescence dots in the flat-mounted retina indi-cated retinal vascular leakage. The fluorescence signal was detected usingan Olympus BX-51 light microscope at ×4 objective (a); scale bar, 100μm. The area (b) and the quantity (c) of the Evans Blue leakage weredetermined (n=6). (d, e) Retinal trypsin digestion was used to detect

changes in the pericytes and the acellular capillaries; scale bar, 25 μm.Representative images are shown (d). Original magnification ×200 usingan Olympus BX-51 light microscope; scale bar, 25 μm. Red arrowsindicate acellular capillaries. Acellular capillaries were quantified in 30random fields per retina and averaged (n=5) (e). Results are presented asmeans ± SEM; *p<0.05 and **p<0.01 for each pair of groups indicated.Con, vehicle control-treated non-diabetic rats

Diabetologia (2019) 62:335–348 339

neovascularisation were potently exacerbated (Fig. 2c–e).Meanwhile, rats pre-treated with AH6809 demonstratedimproved histological retinal changes under diabeticconditions (Fig. 2c–e).

PGE2/EP2R signalling regulates endothelial cell apoptosis inthe retina of STZ-induced diabetic rats In diabetic retinopathy,endothelial cell injury and apoptosis are thought to beone of the initial pathological changes responsible forbreakdown of the blood–retina barrier and the subsequentvascular hyperpermeability [20, 21]. In advanced diabeticretinopathy, there is a more severe loss of endothelial cells

due to retinal hypoperfusion and hypoxia, as well as theaberrant formation of new blood vessels [22]. Three monthsafter diabetes was successfully established in our ratmodel, the PGE2- and butaprost-treated groups displayedsignificantly increased amounts of diabetes-induced cellapoptosis as assessed by TUNEL assay; compared withthese groups, the rats treated with the EP2R antagonistAH6809 showed decreased levels of apoptosis (Fig. 3a,b).Electron-microscopic examination revealed that PGE2

induced apoptotic nuclear condensations in the endothelialcells when compared with cells from the untreated diabeticrats (Fig. 3c). In addition, the endothelium was partially

6 weeks

9 weeks

OC

TF

undu

sO

CT

Fun

dus

3 months

Con STZ STZ+PGE2

STZ+butaprost STZ+AH6809 STZ+FMSO

8

6

4

2

06 weeks 9 weeks

Hyp

erre

flect

ive

dot n

umbe

r

**

***

***

* *

ConSTZSTZ+PGE2STZ+butaprostSTZ+AH6809STZ+DMSO

250

200

150

100

50

0

GC

L–IP

L th

ickn

ess

(μm

)

****

*

****

Con

STZ+PGE 2

STZ+but

apro

st

STZ+AH68

09

STZ+DM

SOSTZ

Con

STZ+PGE 2

STZ+bu

tapr

ost

STZ+AH68

09

STZ+DM

SOSTZ

600

400

Ret

inal

thic

knes

s (μ

m)

200

0

**

**

*

****

Con STZ

STZ STZ+PGE2 STZ+butaprost STZ+AH6809

STZ+PGE2 STZ+butaprost STZ+AH6809 STZ+DMSO

a

c

b

d

e

Fig. 2 PGE2/EP2R signalling influences morphological changes in theeyes of diabetic rats. (a, b) Retinal images at 6 weeks and 9 weeks afterdiabetes was successfully established in the rat model (a). Note the mor-phological changes in the colour fundus images and the OCT images. Thenumber of hyperreflective dots in the OCT images (arrows) was deter-mined (b). (c–e) H&E staining in paraffin sections of rat retinas 3 months

after establishment of the diabetes model (c); scale bar, 25 μm. GCL–IPLand retinal thickness were evaluated in the H&E-stained sections (d, e).The results are presented as means ± SEM; n=6, *p<0.05 and **p<0.01for each pair of groups indicated. Con, vehicle control-treated non-dia-betic rats; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, innerplexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer

340 Diabetologia (2019) 62:335–348

detached from the basal membrane in the PGE2-treatedgroup. At the same time point, the diabetic rats treated withAH6809 showed attenuated cellular injury (Fig. 3c), whichcorrelated with the TUNEL assay data.

PGE2/EP2R signalling is engaged in leucocyte adhesion in thediabetic rat retina Chronic subclinical inflammatory responseis thought to play a critical role in the pathogenesis of diabeticretinopathy [18, 23]. Leucostasis, a main characteristic of

Con

STZ

STZ+PGE2

STZ+butaprost

STZ+AH6809

STZ+DMSO

50

No.

of T

UN

EL-

posi

tive

endo

thel

ial c

ells 40

30

20

10

0

**

*

*

ConSTZ

STZ+PGE 2

STZ+bu

tapr

ost

STZ+AH68

09

STZ+DM

SO

TUNEL

TUNEL

TUNEL

TUNEL

TUNEL

TUNEL

b

a

c

Fig. 3 PGE2/EP2R signalling regulates endothelial cell apoptosis in theretina of STZ-induced diabetic rats. (a) Apoptotic cells in the retinalsections were detected by TUNEL assay. Retinal vessels were counter-stained with isolectin B4 (IB4, green); scale bar, 25 μm. (b)Quantification of TUNEL-positive cells in the outer nuclear layer, innernuclear layer and ganglion cell layer. The results are presented as means ±SEM; n=6, *p<0.05 for each pair of groups indicated. (c) Representative

electron micrographs of retinal vascular endothelial cells in diabetic ratstreated with intravitreal injection of PGE2 or AH6809, showing samplesfrom two control rats; scale bar, 2 μm. Arrows indicate retinal vascularendothelial cells. Con, vehicle control-treated non-diabetic rats; EC, en-dothelial cell; GCL, ganglion cell layer; INL, inner nuclear layer; IPL,inner plexiform layer; nu, nucleus of retinal vascular endothelial cell;ONL, outer nuclear layer; OPL, outer plexiform layer

Diabetologia (2019) 62:335–348 341

inflammation in diabetic retinopathy, is reported to contributeto retinal capillary closure, non-perfusion, capillary dropoutand local ischaemia [24, 25]. In our rat model, adherent retinalleucocytes were labelled in situ with FITC-linked concanava-lin A. Six weeks after diabetes was successfully established,retinal flat mounts were prepared and the adherent leucocyteswere counted in the blood vessels. Compared with the diabeticretina, a 3.6-fold (n = 7, p < 0.01) and 3.3-fold (n = 7, p < 0.01)increase in the number of adherent leucocytes was seen in thePGE2- and butaprost-treated diabetic retinal capillaries, re-spectively (Fig. 4a,b). A 0.7-fold (n = 7, p < 0.01) decreasein the number of adherent leucocytes was seen in the retinasof the AH6809-treated rats vs untreated diabetic rats (Fig.4a,b).This was consistent with the mRNA level of Icam1 inthe retinas (Fig. 4c).

PGE2/EP2R signalling is involved in the activation of theNLRP3 inflammasome in vivo NLRP3 inflammasome activa-tion has been reported in diabetic retinopathy [26]. Its effectormolecule, IL-1β, mediates leucostasis and apoptosis in retinalcapillary endothelial cells [27]. The NLRP3 inflammasome

comprises the cytoplasmic receptor NLRP3, the adaptor mol-ecule ASC and pro-caspase 1. Assembly of the NLRP3inflammasome requires the association of NLRP3 with ASColigomers via homotypic pyrin domain interactions. In ourstudy, the PGE2-treated diabetic rats displayed augmented for-mation of the diabetes-induced NLRP3–ASC complex, whileAH6809 blocked the STZ-related association of NLRP3 withthe ASC oligomers (Fig. 5a). Moreover, this was consistentwith Il-1β (also known as Il1b) and Nlrp3mRNA levels in theretina (Fig. 5b,c).

PGE2/EP2R signalling mediates activation of the NLRP3–ASCcomplex inflammasome in hRMECs Endothelial cells arerecognised as the primary cellular targets for diabetes-induced vascular damage [20, 28, 29]. We selected anRMEC line to study the mechanistic aspects and the function-al significance of PGE2/EP2R signalling alteration in vitro.The association of NLRP3 with the ASC oligomers signifi-cantly increased in the PGE2- and butaprost-treated cells; theassociation was inhibited by AH6809 pre-treatment (Fig. 6a).To further assess the impact of the PGE2/EP2R signalling

Con STZ STZ+PGE2

STZ+butaprost STZ+AH6809 STZ+DMSO

300

200

0

400

100

250

150

100

0

200

50No.

of a

dher

ent

leuc

ocyt

es p

er fi

eld

Icam

1 m

RN

A le

vel (

fold

)

cb*

****

*

**

**

***

a

ConSTZ

STZ+PGE 2

STZ+bu

tapr

ost

STZ+AH68

09

STZ+DM

SOCon

STZ

STZ+PGE 2

STZ+AH68

09

STZ+DM

SO

Fig. 4 PGE2/EP2R signalling isengaged in leucocyte adhesionand inflammation in the retina ofdiabetic rats. (a) Representativeimages of retina sections showingadherent leucocytes labelled withFITC–concanavalin Awithin thediabetic rat retinal vessels; scalebar, 50 μm. (b) Quantification ofadherent leucocytes in the retinasections. (c) Icam1 mRNAexpression in retina (foldnormalised to Gapdh) wasdetermined by real-time PCR.The results are presented asmeans ± SEM; n=6, *p<0.05 and**p<0.01 for each pair of groupsindicated. Con, vehicle control-treated non-diabetic rats

342 Diabetologia (2019) 62:335–348

alteration on inflammasome activation, we examined the mat-uration of pro-caspase 1, which is cleaved into active 10 kDaor 20 kDa fragments that then enzymatically cleave pro-IL-1βto produce mature IL-1β. Treatment with high glucose or

LPS + ATP increased both the caspase 1 cleavage and thelevels of IL-1β and NLRP3 mRNA in hRMECs (ESM Fig.2a,b). A similar effect was seen in hRMECs treated with in-creasing concentrations of PGE2 (1–20 μmol/l) (ESM Fig.

STZ

Con

STZ+PGE2

STZ+AH6809

STZ+DMSO

200

150

100

5040302010

0Il-1β

mR

NA

(fo

ld)

Nlrp

3 m

RN

A (

fold

) 10

6

8

4

2

0

ConSTZ

STZ+PGE 2

STZ+AH68

09

STZ+DM

SOCon

STZ

STZ+PGE 2

STZ+AH68

09

STZ+DM

SO

****

**

***

* *

b c

a

Fig. 5 PGE2/EP2R signalling is involved in the activation of the NLRP3inflammasome in vivo. (a) Representative images of the immunofluores-cence staining of NLRP3 and ASC in the retinal sections of each group ofrats; scale bar, 25 μm. (b, c) Il-1β and Nlrp3 mRNA expression in retina(fold normalised to Gapdh) was measured by real-time PCR. The results

are presented as means ± SEM; n=6, *p<0.05 and **p<0.01 for each pairof groups indicated. Con, vehicle control-treated non-diabetic rats; GCL,ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer;ONL, outer nuclear layer; OPL, outer plexiform layer

Diabetologia (2019) 62:335–348 343

2c). Caspase 1 cleavage in the hRMECs was increased byboth PGE2 and butaprost treatment; the effect was diminishedin the cells pre-treated with AH6809 (Fig. 6b). AH6809 ad-ministration also significantly attenuated PGE2-induced in-creases in the mRNA levels of IL-1β and NLRP3 (Fig. 6c).We next addressed whether IL-1β administration could medi-ate apoptosis of hRMECs. IL-1β (10 ng/ml) stimulation sig-nificantly increased the activation of caspase 3, consistentwith the number of apoptotic cells measured by flow cytom-etry analysis (Fig. 6d,e). As previously mentioned, pyroptosisis initiated by caspase 1 activation and leads tomembrane pore

formation and the leakage of cellular contents. To confirm thatthe cell death induced by PGE2 and butaprost was pyroptosis,the caspase 1 inhibitor, z-YVAD-fmk (z-VAD), was used. z-VAD pre-treatment effectively abolished the LDH release andthe caspase 3 activation induced by PGE2 and butaprost ex-posure (Fig. 6f,g).

EP2R-coupled cAMP/protein kinase A/cAMP responseelement-binding protein signalling mediated NLRP3 activa-tion and pyroptosis in RMECs We next examined the signal-ling pathways involved in regulating the PGE2/EP2R-

Con

IP: NLRP3

IB: ASC

IB: NLRP3

IB: Tubulin

PGE2PGE2+AH6809

42K

Rel

ativ

e bi

ndin

g in

tens

ity(N

LRP

3-A

SC

) (f

old)

5

*****

**4

3

2

1

0

110K

51K

Butaprost

Con

PGE 2

Butap

rost

PGE 2+AH68

09

Con PGE2 ButaprostPGE2+AH6809

Caspase 1

Cleavedcaspase 1 P20

β-Actin

σ

γ

β

α

Cle

aved

cas

pase

1/

pro

casp

ase

1 (f

old)

3

****

* **

****

**

**

2

1

0

****

Con

PGE 2

Butap

rost

PGE 2+AH68

09

3

2

1

0

Rel

ativ

e m

RN

A le

vel

(fol

d)

**ConPGE2ButaprostPGE2+AH6809

IL-1β NLRP3

IL-1β (ng/ml)

β-Actin

Cleavedcaspase 3

0 0.1 10

Cle

aved

cas

pase

3/

IL-1

β (n

g/m

l) pr

o-ca

spas

e 3

(fol

d)

1.6

1.4

1.2

1.0

0.80 1 100.1

100

101

102

PI

AV

103

104

100 101 102 103 104

Con

5.32±1.21%

Caspase 3

1

IL-1β

8.61±1.21%

100 101 102 103 104

AV

101

100

102

103

104

PI

Con IL-1β

**15

10

5

0Cel

l apo

ptos

is (

%)

****

** *6

4

2

0

LDH

act

ivity

(U

/gpr

ot)

Con

PGE 2

Butap

rost

Butap

rost+z-

VAD

PGE 2+z-

VAD

ConPGE 2

Butap

rost

Butap

rost+z-

VAD

PGE 2+z-

VAD

Con PGE2 ButaPGE2+z-VAD

Buta+z-VAD

Caspase 3

Cleavedcaspase 3

β-Actin

2.5

2.0

1.5

1.0

0.5

0

Cle

aved

cas

pase

3/

pro

casp

ase

3 (f

old)

**

** **

a

c

e

g

f

d

b

Fig. 6 PGE2/EP2R signalling mediates the activation of the NLRP3–ASC complex in hRMECs. The hRMECs were treated with 1 μmol/lPGE2 for 24 h, with or without 30 min AH6809 pre-treatment. (a)Immunoprecipitation was performedwith anti-NLRP3 antibody followedby immunoblotting with ASC antibody; n=5. (b) The levels of pro-cas-pase 1 and cleaved caspase 1 (P20) were determined bywestern blot; n=5.(c) NLRP3 and IL-1βmRNA expression in hRMECs (fold normalised toGAPDH) was detected by real-time PCR; n=4. (d, g) The levels of

caspase 3 and cleaved caspase 3 in response to different treatments asindicated were determined by western blot; n=5. (e) Apoptosis of thehRMECs was determined by annexin V–FITC/propidium iodide flowcytometry; n=6. (f) LDH activity was measured in each group; n=4.Representative blots are shown, with quantification. Data are presentedas means ± SEM. *p<0.05 and **p<0.01 for each pair of groups indicat-ed. AV, annexin V; Buta, butaprost; Con, vehicle control; gprot, gramprotein; PI, propidium iodide

344 Diabetologia (2019) 62:335–348

mediated NLRP3 activation and pyroptosis in hRMECs.Recent studies have shown that PGE2 activates EP2R, stimu-lating cAMP/protein kinase A (PKA) signalling [30, 31].Indeed, in our study, pre-treatment of the hRMECs with eitherSQ22536 (an adenylate cyclase inhibitor) or H89 (a PKAinhibitor) dramatically inhibited the PGE2-induced NLRP3activation (Fig. 7a,b). This was consistent with the decreasedIL-1β and NLRP3mRNA levels, as well as the attenuated IL-1β production (Fig. 7c,d). Epac1 and PKA are both intracel-lular receptors of cAMP. However, Epac1 expression was sig-nificantly inhibited by high glucose (30 mmol/l) comparedwith normal glucose (5 mmol/l) treatment (ESM Fig. 3a,b).Epac1 agonist had no effect on PGE2-induced NLRP3inflammasome activation in hRMECs (ESM Fig. 3c,d).Under the circumstances, diabetic retinopathy and high-glucose stress-associated PGE2/EP2R-mediated NLRP3 acti-vation may mainly result from the cAMP/PKA/cAMP re-sponse element-binding protein (CREB) signalling pathway.It has been reported that activated PKA transfers into the cell

nucleus and phosphorylates the transcription factor CREBprotein, regulating gene expression. Therefore, we exam-ined the phosphorylation of CREB at Ser133, which is crit-ical for CREB transcriptional activation. Stimulation ofhRMECs with PGE2 and butaprost led to a significant in-crease in CREB phosphorylation, which was blocked byAH6809 pre-treatment (Fig. 7e). PGE2 induced the phos-phorylation and activation of CREB in rats and this was alsoblocked by AH6809 pre-treatment (ESM Fig. 4a,b).We alsoinvestigated the effects of AH6809 on the activation ofNF-κB, another key transcription factor implicated inPGE2/EP2R signalling pathways in hRMECs. There wasno significant difference noted between each group ofhRMECs (ESM Fig. 5a,b). Moreover, both the adenylatecyclase inhibitor (SQ22536) and the PKA inhibitor (H89)inhibited CREB phosphorylation (Fig. 7f). Taken together,this data suggests that the cAMP/PKA/CREB signallingpathway is involved in EP2R-mediated NLRP3 activationand pyroptosis in hRMECs.

Con PGE2Buta-prost

PGE2+

AH6809PGE2

+H89

PGE2+

SQ22536Con PGE2

Buta-prost

PGE2+

AH6809PGE2

+H89

PGE2+

SQ22536

Con PGE2

Buta-prost

PGE2+

AH809PGE2

+H89

PGE2+

SQ22536PGE2

+DMSO

Cytoplasm

PGE2Con Con PGE2Buta-prost

Buta-prost

PGE2+

AH809PGE2

+AH6809

IP: NLRP3

IB: ASC

IB:NLRP3

IB:β-Actin

****

****

**5

4

3

2

1

0

Rel

ativ

e bi

ndin

g in

tens

ity(N

LRP

3-A

SC

) (f

old)

ConPGE 2

PGE 2+AH68

09

PGE 2+H89

PGE 2+SQ22

536

Butap

rost

ConPGE 2

PGE 2+H89

PGE 2+SQ22

536

ConPGE 2

Butap

rost

PGE 2+AH68

09 ConPGE 2

Butap

rost

PGE 2+AH68

09

PGE 2+H89

PGE 2+SQ22

536

PGE 2+DM

SO

ConPGE 2

PGE 2+AH68

09

PGE 2+H89

PGE 2+SQ22

536

Butap

rost

Caspase 1

Cleavedcaspase 1

β-Actin

****

***

*2.0

1.5

1.0

0.5

0

Cle

aved

cas

pase

1/

pro-

casp

ase

1 (f

old)

3

2

1

0

Rel

ativ

e m

RN

A le

vel

(fol

d)

βIL-1 NLRP3

****** **

*** PGE2+H-89

PGE2+SQ22536

PGE2

Con** *

*

IL-1

β (p

g/m

g pr

otei

n)

2000

1500

1000

500

0p-CREB

CREB

Laminb

p-CREB

CREB

Laminb

β-Actin

Nucleus

**

**

****

***

* **

*2.5

2.0

1.5

1.0

0.5

0

p-C

RE

B/C

RE

B (

fold

)

p-C

RE

B/C

RE

B (

fold

)

4

3

2

1

0

a

c

e

d

b

f

Fig. 7 EP2R-coupled cAMP/PKA/CREB signalling mediates NLRP3activation and pyroptosis in hRMECs. (a) Immunoprecipitation was per-formed with anti-NLRP3 antibody followed by immunoblotting withASC antibody; n=4. (b) The levels of pro-caspase 1 and cleaved caspase1 (P20) were determined by western blot; n=5. (c) NLRP3 and IL-1βmRNA expression in hRMECs (fold normalised to GAPDH) was deter-mined by real-time PCR; n=4. (d) IL-1β was detected in the cell culture

supernatant fractions of hRMECs by ELISA; n=3. (e, f) The levels of p-CREB and CREB were determined by western blot. Laminb antibodywas used to confirm equal nuclear protein loading among samples; n=5.Representative blots are shown, with quantification. The results are pre-sented as means ± SEM. *p<0.05 and **p<0.01 for each pair of groupsindicated. Con, vehicle control

Diabetologia (2019) 62:335–348 345

Discussion

In the present study, we demonstrated the important role ofPGE2/EP2R signalling in diabetes-associated hRMEC dys-function and vascular dysfunction. We found that thediabetes-associated inflammation and hRMEC damage wereameliorated by inhibition of the PGE2/EP2R and coupledcAMP–PKA–CREB signalling pathways. This was associat-ed with an amelioration of NLRP3 inflammasome activation.

Animals in the STZ-induced diabetes model displayedsymptoms of diabetic retinopathy such as endothelium impair-ment, vascular leakage and capillary degeneration, leading toincreased acellular vessels [32–34]. We showed that the ret-inas of STZ-induced diabetic rats developed phenotypical andhistopathological features consistent with diabetic retinopathy.PGE2, butaprost, AH6809 and DMSO were administered tothe rats by intravitreal injection, suggesting that their retinaleffects are independent of any systemic activity. We observedthat both PGE2 and butaprost aggravated the deleterious ef-fects of STZ-induced diabetes, while the effects were amelio-rated in the animal group that received AH6809.

OCT is a non-invasive imagingmodality that enables quan-titative measurement of retinal thickness and evaluation ofmorphological changes in eyes with diabetic retinopathy anddiabetic macular oedema [35]. Consistent with the histologicalexamination, the OCT images revealed retinal oedema and anincreased number of hyperreflective dots in the inner retina ofPGE2- and butaprost-treated diabetic rats. The hyperreflectivedots in diabetic retina delineated on spectral domain OCTrepresent activated microglia cells and it has been reportedthat the number increases with progressing retinopathy [36,37]. These hyperreflective dots have been described in inflam-matory retinal conditions and may present in diabetic eyeseven when clinical retinopathy is undetectable [38, 39].Therefore, hyperreflective dots in OCT imaging of diabeticretinas are a prominent feature of the disease process andmay be used to closely monitor diabetic retinopathy in clinicalpractice.

In the early stages of diabetic retinopathy, hyperglycaemiaand chronic inflammation damage the retinal endothelium andplay a key role in further vascular leakage, pericyte loss, in-creased acellular vessels and the eventual manifestation ofclinical diabetic retinopathy symptoms [40, 41]. Leucostasisis a characteristic of diabetic retinopathy inflammation [25,42]. Our results demonstrate that PGE2 and butaprost treat-ment of hRMECs increases diabetes-induced leucocyte infil-tration, Icam1 expression and apoptosis of RMECs. Theseeffects were suppressed by AH6809 pre-treatment. However,the sample size of the AH6809 + STZ group in our study wasnot large enough to draw an irrefutable conclusion regardingIcam1 expression.

Recent studies have implicated the inappropriate activationof the NLRP3 inflammasome in diabetic retinopathy [26].

Glucose abnormalities have also been reported to be an im-portant trigger of the sterile inflammatory response mediatedby the NLRP3 inflammasome [43]. We observed that PGE2

and butaprost promoted diabetes-induced activation ofNLRP3 inflammasome in vivo and in vitro and that this couldbe prevented by AH6809 administration. Moreover, exoge-nous administration of IL-1β led to an elevation in hRMECapoptosis. Alternatively, however, it was also reported thatPGE2 inhibits NLRP3 inflammasome activation throughEP4R and intracellular cAMP in human macrophages [44].This difference might be due to the dose and the way in whichPGE2 was used in those studies (PGE2 was used after LPSpriming). The differences in our findingsmight also arise fromcell type and epoprostanoid expression. This data suggeststhat the PGE2/EP2R signalling pathway mediates NLRP3inflammasome activation in diabetic retinopathy.

The G protein consists of α, β and γ subunits, and the αsubunit is divided into Gαs, Gαi and Gαq, among others.According to previous reports, EP2R couples to the Gαs sub-unit, activates adenylate cyclase, increases cytoplasmic cAMPand induces PKA activation [12]. Our data showed that thePKA inhibitor H89 and the adenylate cyclase inhibitorSQ22536 suppressed the increase in IL-1β and NLRP3mRNA levels, as well as the increase in IL-1β levels, inducedby PGE2. This demonstrates that cAMP and PKA are in-volved in the signalling pathway mediated by PGE2/EP2R.

The transcription factors CREB and NF-κB play pivotalroles in the PGE2 signalling pathway and the developmentof diabetic retinopathy [45–48]. Toll-like receptor-mediatedNF-κB activation is associated with an acute activation ofthe NLRP3 inflammasome and production of IL-1β in mac-rophages. After these acute and initial signals are received,adenosine further regulates IL-1β production by activatingthe cAMP–PKA–CREB signalling cascade, resulting in theupregulation of pro-IL-1β and NLRP3, further activating cas-pase 1, without the need for any other initiating signals [49,50]. This signalling pathway has an established important rolein several chronic inflammatory diseases. Likewise, we ob-served in hRMECs that the ability of the PGE2/EP2R–cAMP–PKA signalling pathway to upregulate pro-IL-1βwas dependent on CREB activation, while there was no sig-nificant difference in the activation of NF-κB.

In summary, we demonstrated that disruption of the PGE2/EP2R signalling pathway contributes to the attenuation of di-abetic retinopathy. The underlying mechanism is multifold.First, long-term exposure to high glucose concentrations andother diabetes risk factors increases the expression and activa-tion of COX-2, subsequently promoting the PGE2/EP2R–cAMP–PKA signalling pathway. PKA transfers into the cellnucleus and phosphorylates the transcription factor CREB,upregulating the transcription of NLRP3 and pro-IL-1β.Although the molecular basis for the activation of the PGE2/EP2R cascade in diabetic retinopathy remains to be delineated,

346 Diabetologia (2019) 62:335–348

the present study implies that the PGE2/EP2R signalling path-way is a target for new therapeutic strategies to prevent andtreat diabetic retinopathy.

Data availability All relevant data are included in the article and/or theESM files.

Funding This work was supported by grants from the National NaturalScience Foundation of China (no. 81770941), Jiangsu Key MedicalDisciplines (ZDXKC2016008), Technology Development Fund(CSE12N1701) and Wuxi Eminent Medical Talents (JCRC005).

Duality of interest The authors declare that there is no duality of interestassociated with this manuscript.

Contribution statement XW and YY made substantial contributions toconception and design, acquisition of data, analysis and interpretation ofdata, drafting the article and revising it critically for important intellectualcontent. MW, YW, TX and PZ made substantial contributions to acqui-sition of data, analysis and interpretation of data and revising the articlecritically for important intellectual content. JZ, XN, JLi, YL, QS andJLeng made substantial contributions to conception and design, analysisand interpretation of data and revising the article critically for importantintellectual content. JShao, MZ and CT made substantial contributions toconception and design, acquisition of data and revising the article criti-cally for important intellectual content. JT, YD and JSunmade substantialcontributions to conception and design, analysis and interpretation of dataand revising the manuscript critically for important intellectual content.All authors gave final approval of the version to be published. XW hadfull access to all the data in this study and takes responsibility for theintegrity of the data and the accuracy of the data analysis.

References

1. Zheng Y, He M, Congdon N (2012) The worldwide epidemic ofdiabetic retinopathy. Indian J Ophthalmol 60(5):428–431. https://doi.org/10.4103/0301-4738.100542

2. Duh EJ, Sun JK, Stitt AW (2017) Diabetic retinopathy: currentunderstanding, mechanisms, and treatment strategies. JCI Insight2(14). https://doi.org/10.1172/jci.insight.93751

3. Li AF, Roy S (2009) High glucose-induced downregulation ofconnexin 43 expression promotes apoptosis in microvascular endo-thelial cells. Invest Ophthalmol Vis Sci 50(3):1400–1407. https://doi.org/10.1167/iovs.07-1519

4. Monaghan K, McNaughten J, McGahon MK et al (2015)Hyperglycemia and diabetes downregulate the functional expres-sion of TRPV4 channels in retinal microvascular endothelium.PLoS One 10:e128359

5. Zhang Y, Lv X, Hu Z et al (2017) Protection of Mcc950 againsthigh-glucose-induced human retinal endothelial cell dysfunction.Cell Death Dis 8(7):e2941. https://doi.org/10.1038/cddis.2017.308

6. Liu K, Liu H, Zhang Z, Ye W, Xu X (2016) The role of N-glycosylation in high glucose-induced upregulation of intercellularadhesion molecule-1 on bovine retinal endothelial cells. ActaOphthalmol 94(4):353–357. https://doi.org/10.1111/aos.13028

7. Sutterwala FS, Haasken S, Cassel SL (2014)Mechanism of NLRP3inflammasome activation. Ann N Y Acad Sci 1319(1):82–95.https://doi.org/10.1111/nyas.12458

8. Chen W, Zhao M, Zhao S et al (2017) Activation of the TXNIP/NLRP3 inflammasome pathway contributes to inflammation in di-abetic retinopathy: a novel inhibitory effect of minocycline.Inflamm Res 66(2):157–166. https://doi.org/10.1007/s00011-016-1002-6

9. Zhang Y, Yang X, Qiu C, Liu F, Liu P, Liu Z (2018) Matrinesuppresses AGE-induced HAEC injury by inhibiting ROS-mediated NRLP3 inflammasome activation. Eur J Pharmacol 822:207–211. https://doi.org/10.1016/j.ejphar.2018.01.029

10. Zaslona Z, Palsson-McDermott EM, Menon D et al (2017) Theinduction of pro-IL-1beta by lipopolysaccharide requires endoge-nous prostaglandin E2 production. J Immunol 198(9):3558–3564.https://doi.org/10.4049/jimmunol.1602072

11. Hata AN, Breyer RM (2004) Pharmacology and signaling of pros-taglandin receptors: multiple roles in inflammation and immunemodulation. Pharmacol Ther 103(2):147–166. https://doi.org/10.1016/j.pharmthera.2004.06.003

12. Wu T (2005) Cyclooxygenase-2 and prostaglandin signaling incholangiocarcinoma. Biochim Biophys Acta 1755(2):135–150.https://doi.org/10.1016/j.bbcan.2005.04.002

13. Schoenberger SD, Kim SJ, Sheng J, Rezaei KA, Lalezary M,Cherney E (2012) Increased prostaglandin E2 (PGE2) levels inproliferative diabetic retinopathy, and correlation with VEGF andinflammatory cytokines. Invest Ophthalmol Vis Sci 53(9):5906–5911. https://doi.org/10.1167/iovs.12-10410

14. Kern TS, Miller CM, Du Y et al (2007) Topical administration ofnepafenac inhibits diabetes-induced retinal microvascular diseaseand underlying abnormalities of retinal metabolism and physiology.Diabetes 56(2):373–379. https://doi.org/10.2337/db05-1621

15. Ayalasomayajula SP, Kompella UB (2003) Celecoxib, a selectivecyclooxygenase-2 inhibitor, inhibits retinal vascular endothelialgrowth factor expression and vascular leakage in a streptozotocin-induced diabetic rat model. Eur J Pharmacol 458(3):283–289.https://doi.org/10.1016/S0014-2999(02)02793-0

16. Shan K, Liu C, Liu BH et al (2017) Circular noncoding RNAHIPK3 mediates retinal vascular dysfunction in diabetes mellitus.Circulation 136(17):1629–1642. https://doi.org/10.1161/CIRCULATIONAHA.117.029004

17. Wang X, Gao Y, Song J et al (2017) The TIR/BB-loopmimetic AS-1 prevents non-alcoholic steatohepatitis and hepatic insulin resis-tance by inhibiting NLRP3-ASC inflammasome activation. Br JPharmacol 174(12):1841–1856. https://doi.org/10.1111/bph.13786

18. Noda K, Nakao S, Zandi S, Sun D, Hayes KC, Hafezi-MoghadamA (2014) Retinopathy in a novel model of metabolic syndrome andtype 2 diabetes: new insight on the inflammatory paradigm. FASEBJ 28(5):2038–2046. https://doi.org/10.1096/fj.12-215715

19. Zhang J, Wu Y, Jin Y et al (2008) Intravitreal injection of erythro-poietin protects both retinal vascular and neuronal cells in earlydiabetes. Invest Ophthalmol Vis Sci 49(2):732–742. https://doi.org/10.1167/iovs.07-0721

20. Watson EC, Grant ZL, Coultas L (2017) Endothelial cell apoptosisin angiogenesis and vessel regression. Cell Mol Life Sci 74(24):4387–4403. https://doi.org/10.1007/s00018-017-2577-y

21. Feenstra DJ, Yego EC,Mohr S (2013)Modes of retinal cell death indiabetic retinopathy. J Clin Exp Ophthalmol 4:298

22. Antonetti DA, Klein R, Gardner TW (2012) Diabetic retinopathy. NEngl J Med 366(13):1227–1239. https://doi.org/10.1056/NEJMra1005073

23. Wu H, Hwang DK, Song X, Tao Y (2017) Association betweenaqueous cytokines and diabetic retinopathy stage. J Ophthalmol2017:9402198

24. Joussen AM, Murata T, Tsujikawa A, Kirchhof B, Bursell SE,Adamis AP (2001) Leukocyte-mediated endothelial cell injuryand death in the diabetic retina. Am J Pathol 158(1):147–152.https://doi.org/10.1016/S0002-9440(10)63952-1

25. Rangasamy S, McGuire PG, Franco NC, Monickaraj F, OrugantiSR, Das A (2014) Chemokine mediated monocyte trafficking intothe retina: role of inflammation in alteration of the blood-retinalbarrier in diabetic retinopathy. PLoS One 9(10):e108508. https://doi.org/10.1371/journal.pone.0108508

Diabetologia (2019) 62:335–348 347

26. Chen H, Zhang X, Liao N et al (2018) Enhanced expression ofNLRP3 inflammasome-related inflammation in diabetic retinopa-thy. Invest Ophthalmol Vis Sci 59(2):978–985. https://doi.org/10.1167/iovs.17-22816

27. Zhu X, Xie M, Wang K et al (2014) The effect of puerarin againstIL-1beta-mediated leukostasis and apoptosis in retinal capillary en-dothelial cells (TR-iBRB2). Mol Vis 20:1815–1823

28. Zhang W, Chen S, Liu ML (2018) Pathogenic roles ofmicrovesicles in diabetic retinopathy. Acta Pharmacol Sin 39(1):1–11. https://doi.org/10.1038/aps.2017.77

29. Eshaq RS, Aldalati A, Alexander JS, Harris NR (2017) Diabeticretinopathy: breaking the barrier. Pathophysiology 24(4):229–241.https://doi.org/10.1016/j.pathophys.2017.07.001

30. Brudvik KW, Tasken K (2012) Modulation of T cell immune func-tions by the prostaglandin E(2) - cAMP pathway in chronic inflam-matory states. Br J Pharmacol 166(2):411–419. https://doi.org/10.1111/j.1476-5381.2011.01800.x

31. Luo X, Zhu Q, Zhang J, Huang Q, Xie Z, Cheng Y (2017) Thedouble roles of the prostaglandin E2 EP2 receptor in intracerebralhemorrhage. Curr Drug Targets 18(12):1377–1385. https://doi.org/10.2174/1389450117666151209122826

32. Olivares AM, Althoff K, Chen GF et al (2017) Animal Models ofDiabetic Retinopathy. Curr Diab Rep 17(10):93. https://doi.org/10.1007/s11892-017-0913-0

33. Lai AK, Lo AC (2013) Animal models of diabetic retinopathy:summary and comparison. J Diabetes Res 2013:106594

34. Fong DS, Aiello L, Gardner TW et al (2003) Diabetic retinopathy.Diabetes Care 26(1):226–229. https://doi.org/10.2337/diacare.26.1.226

35. Sun JK, Lin MM, Lammer J et al (2014) Disorganization of theretinal inner layers as a predictor of visual acuity in eyes withcenter-involved diabetic macular edema. JAMA Ophthalmol132(11):1309–1316. https://doi.org/10.1001/jamaophthalmol.2014.2350

36. Vujosevic S, Bini S, Midena G, Berton M, Pilotto E, Midena E(2013) Hyperreflective intraretinal spots in diabetics without andwith nonproliferative diabetic retinopathy: an in vivo study usingspectral domain OCT. J Diabetes Res 2013:491835

37. Garner A (1981) Developments in the pathology of diabetic reti-nopathy: a review. J R Soc Med 74(6):427–431. https://doi.org/10.1177/014107688107400607

38. Zeng HY, Green WR, Tso MO (2008) Microglial activation inhuman diabetic retinopathy. Arch Ophthalmol 126(2):227–232.https://doi.org/10.1001/archophthalmol.2007.65

39. Ogino K, Murakami T, Tsujikawa A et al (2012) Characteristics ofoptical coherence tomographic hyperreflective foci in retinal veinocclusion. Retina 32(1):77–85. https://doi.org/10.1097/IAE.0b013e318217ffc7

40. Dejana E, Tournier-Lasserve E, Weinstein BM (2009) The controlof vascular integrity by endothelial cell junctions: molecular basisand pathological implications. Dev Cell 16(2):209–221. https://doi.org/10.1016/j.devcel.2009.01.004

41. Ogura S, Kurata K, Hattori Y et al (2017) Sustained inflammationafter pericyte depletion induces irreversible blood-retina barrierbreakdown. JCI Insight 2:e90905

42. Vestweber D (2015) How leukocytes cross the vascular endotheli-um. Nat Rev Immunol 15(11):692–704. https://doi.org/10.1038/nri3908

43. Zhang Y, Lv X, Hu Z et al (2017) Protection of Mcc950 againsthigh-glucose-induced human retinal endothelial cell dysfunction.Cell Death Dis 8(7):e2941. https://doi.org/10.1038/cddis.2017.308

44. Sokolowska M, Chen LY, Liu Y et al (2015) Prostaglandin E2inhibits NLRP3 inflammasome activation through EP4 receptorand intracellular cyclic AMP in human macrophages. J Immunol194(11):5472–5487. https://doi.org/10.4049/jimmunol.1401343

45. Devi TS, Singh LP, Hosoya K, Terasaki T (2011) GSK-3β/CREBaxis mediates IGF-1-induced ECM/adhesion molecule expression,cell cycle progression and monolayer permeability in retinal capil-lary endothelial cells: Implications for diabetic retinopathy.Biochim Biophys Acta 1812(9):1080–1088. https://doi.org/10.1016/j.bbadis.2011.04.007

46. Graves DT, Kayal RA (2008) Diabetic complications and dysregu-lated innate immunity. Front Biosci 13(13):1227–1239. https://doi.org/10.2741/2757

47. Luan B, YoonYS, Le Lay J, Kaestner KH, Hedrick S,MontminyM(2015) CREB pathway links PGE2 signaling with macrophage po-larization. Proc Natl Acad Sci U S A 112(51):15642–15647. https://doi.org/10.1073/pnas.1519644112

48. Steer SA, Corbett JA (2003) The role and regulation of COX-2during viral infection. Viral Immunol 16(4):447–460. https://doi.org/10.1089/088282403771926283

49. Baron L, Gombault A, Fanny M et al (2015) The NLRP3inflammasome is activated by nanoparticles through ATP, ADPand adenosine. Cell Death Dis 6(2):e1629. https://doi.org/10.1038/cddis.2014.576

50. Ouyang X, Ghani A,Malik A et al (2013) Adenosine is required forsustained inflammasome activation via the A2A receptor and theHIF-1α pathway. Nat Commun 4(1):2909. https://doi.org/10.1038/ncomms3909

348 Diabetologia (2019) 62:335–348